Protocol for a prospective observational cohort study of cutaneous leishmaniasis in Ethiopia

Objectives

Patient and Public Involvement

Patients with CL were involved in the validation of the translations of PROMs through focus group discussions, cultural adaptation and piloting.

Study context

The study will enrol individuals receiving treatment for CL at two centres: ALERT Hospital, Addis Ababa and Boru Meda Hospital, South Wollo Zone, Amhara Region. ALERT Hospital serves as both a general hospital for patients in Addis Ababa and a regional and national specialized referral centre for individuals with CL and other skin conditions. The hospital has nine dermatologists and a 20-bed unit for the care of patients with CL and diagnoses approximately 1000 cases of CL managed on an inpatient or outpatient basis each year. Boru Meda serves as the general hospital and specialised referral service for all skin conditions in South Wollo zone. It has two dermatologists, and a 35-bed unit for the management of CL and diagnoses around 700 cases of CL, managed on an inpatient or outpatient basis each year.

Timelines

Study recruitment began in April 2022 and is expected to continue until the sample size is reached or until the cut-off of September 30th, 2023, whichever occurs first. Follow-up is therefore expected to be completed by September 30th, 2024.

Study participants and management

All adults and children diagnosed with CL and initiated on CL treatment at both study centres are eligible to be recruited in this study. Each participant will be followed for 360 days.

Inclusion criteria:

1. Individuals with parasitologically confirmed CL using microscopy or culture

2. Individuals receiving treatment at the study centres

Exclusion criteria:

1. Participants who have previously been treated with systemic anti-leishmania therapy

2. Participants who do not or are unable to give informed consent to participate in the study

Participants will be treated using the standard of care for CL at the centre where they are managed and in accordance with the Ethiopian national guideline²⁴ which will be agreed by two clinicians in consultation with the affected individual prior to enrolment. The treatment may be changed if deemed clinically appropriate. In general, if the size of the skin lesion is 4cm in diameter of smaller and not near a mucosal site the first line treatment is most frequently local therapy (cryotherapy and/or intralesional SSG). In general, if the diameter of the skin lesion is greater than 4 cm and/or near or involving a mucosal surface the initial therapy is usually intramuscular SSG for 28 days with or without adjunctive cryotherapy. Participants will be assessed on day 42 and further local or systemic therapy may be offered depending on the degree of clinical response.

We will classify CL using the operational definitions shown in Box 1.

Overview of data collection

This prospective observational cohort will integrate different methodologies appropriate to the different “areas” under investigation, using qualitative and quantitative tools:

Study visits for data collection are scheduled at day 1, 14, 28, 42, 90, 180 and 360 (Table 1). Treatment will be decided by the treating clinicians, who may or may not be one of the study dermatologists. All study specific data collection will be undertaken by a member of the study team. At the baseline visit, we will collect socio-demographic data, perform clinical assessment, and collect data on PROMs and costs. Follow-up data will be collected for all patients at specified time points, as detailed further below. In a subset of patients, additional biological samples will be collected for immunological and genetic studies. A further subset of participants will be selected for in-depth interviews.

Clinical data

At the baseline visit, we will record the clinical phenotype of CL including the number, size, morphology²⁵ and location of lesions. Lesion dimensions (defined as the largest diameter of the lesion) will be measured using a disposable tape measure. Individuals with more than one lesion will be asked to indicate the lesion causing greatest concern and this will be designated the “index” lesion. Individuals may indicate two or more lesions of equal concern. In those with mucosal involvement, mucosal signs and symptoms are recorded. Clinicians will then provide a subjective global assessment of CL severity using a 3-point Likert scale (mild, moderate, severe). Slit-skin smears will be collected at baseline for microscopy and species identification using polymerase chain reaction (PCR).

At each follow-up visit, we will assess lesions for the degree of re-epithelization or flattening. For those with DCL, the number of resolved lesions and percent of lesions flattened per region is recorded. Clinicians will provide a global assessment of the improvement or worsening of CL using a 5-point Likert scale (worse, no change, minor improvement, major improvement, cure). At each visit, we will record the treatment being administered and any changes since the last visit. The adverse reactions to CL treatments will be assessed including the nature and severity of the reaction, the action taken, the presumed cause and the outcome. The Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 will be used for describing adverse reactions and grading their severity. Slit-skin smear samples will be collected at day 90 for test of cure if the skin lesion is not healed.

PROMs

At baseline, Amharic or Oromiffa translations of several PROMs will be collected (Table 1): two general measures of HRQoL, the EQ-5D-5L and the SF-36, and two skin-specific measures of HRQoL, the Dermatology Life Quality Index (DLQI) and the Children’s Dermatology Life Quality Index (CDLQI). Participants will complete the Cutaneous Leishmaniasis Impact Questionnaire (CLIQ), a CL-specific measure of HRQoL.

At the follow-up visits indicated in Table 1, we will collect PROMs again and ask participants for their global assessment of the improvement of their CL lesion(s) using a 5-point Likert scale (better, somewhat better, same, somewhat worse, worse). The Amharic or Afan Oromo translations of the Patient and Observer Scar Assessment Scale (POSAS) v2²⁶ will be completed by both the clinician and participant at the day 42 assessment and each subsequent visit.

Lived experience

A selection of participants will be recruited to participate in an in-depth interview to capture data on their lived experience of CL until data saturation has occurred. Initially, we aim to recruit a total of 30 adults and 10 children from both study centres. We will ensure representation of participants by key demographics including age, gender, and CL phenotype. Interviews will be conducted by members of the research team using a topic-guide covering key themes including the impact of the illness, the care-seeking pathways they have undertaken and their experience of traditional and allopathic medicine. Ethnographic observation will be used to enrich the data on treatment experiences in the health facilities.

Costs

At baseline data will be collected on costs incurred by the individuals and their household that can be attributed to a member having CL and/or seeking treatment for CL including costs incurred since onset of CL symptoms. We will also collect data on household income, expenditure, and asset possession to establish a month of income and expenditure data.

At each follow up visit, cost related to CL since the last visit/admission will be collected. These will include costs of medicines, hospitalisation, consultation, transportation, and opportunity costs related to having and treating CL (i.e., missed work or school). We will also collect information on how households cope with the economic burden of CL.

Data from the provider perspective will be collected. This will include costs of training health workers, providing services, medicines donated, and other service-related costs. These costs will be collected from the project’s activity and financial records; the value of any donated items will be estimated based on market prices. Where necessary, government and donor sources will be consulted for relevant cost data.

Immunology samples

To interrogate the dynamics of immune responses in different clinical phenotypes of CL, blood and skin lesion samples will be collected at three time points. Participants will have blood samples and a skin biopsy (stored in RNAlater™) taken to evaluate subsets of the innate and adaptive immunity; pro-inflammatory (IFN-γ and TNF-α) and anti-inflammatory (IL-6, IL-27, IL-10, IL-4, IL-13, TGF-β, and hsp18) cytokines and interleukins (IL)-12, IL-2, IL-1, IL-18, IL-15, IL-8, IL-17A, and IL-22) using Luminex multiplex assay. (ii) The change in mRNA (Ribonucleic acid) expression level of Toll-like receptors (TLRs) and pro-inflammatory (IFN-γ, IL-12, IL-1β, IL-17A, and IL-12p70) and anti-inflammatory (IL-5, IL-6, IL-10, IL-4, TGF-β, and hsp18) cytokines will also be assessed before treatment (at day 0), during treatment (day 28) and after completion of treatment (day 90) using quantitative PCR (Real-time Polymerase Chain Reaction). In addition, the levels of chemokines (such as CXCL8, CXCL10, CCL5, CCL4, CCL3, and CCL2) will be determined in supernatants from the whole blood stimulation assays for the Luminex multiplex assay. HLA typing will be performed to assess if there is an association between HLA types (targeting probes for HLA-A, B, C, DR, DQ, and DP loci) and the clinical phenotypes of CL seen in our cohort.

Sample Size

The primary outcome is the proportion of individuals with a healed index lesion by study day 90. In line with the consensus document on clinical trial outcomes cure will be defined as 100% re-epithelialization or flattening of the index lesion. Pre-specified additional outcomes are the proportion healed at day 42 and 180. Secondary outcomes include the frequency of adverse events and measures of HRQoL life.

There are no pre-existing high-quality data on cure rates for CL caused by L. aethiopica. We calculated the sample size required to measure a range of cure-rates at 90 days with varying degrees of precision (Table 2). Based on this we aimed to recruit up to 250 participants with each CL phenotype (LCL, MCL, and DCL) across the two sites. We anticipate that mucosal leishmaniasis will be rare and will recruit as many individuals with this type of disease as possible.

Data analysis

Objective 1: To assess clinical outcomes in LCL, MCL and DCL in a prospective cohort. & Objective 2: To assess clinical parameters in LCL, MCL and DCL in a prospective cohort.

We will use descriptive statistics (means, medians, and proportions) to describe demographics and baseline characteristics of the cohort. In exploratory analysis, we will use univariable logistic regression to examine clinical and demographic variables associated with resolution of CL at Day 42, 90, and 180. Whilst not formally powered for these analyses, the findings will be viewed as hypothesis-generating and potentially form the basis for future studies.

Objective 3: To assess the change in HRQoL associated with CL before and after treatment.

Changes in general, skin and CL-specific HRQoL will be assessed using the following PROMs: EQ-5D-5L, SF-36, DLQI/ CDLQI and CLIQ applied at the timepoints shown in Table 1, and compared with clinical outcomes to determine validity and responsiveness.

Validity and responsiveness of each PROM will be analysed using SPSS version 23.0 (IBM Corporation, Armonk, NY, USA). Descriptive statistics will be presented using frequencies, percentages, tables, and graphs.

Mean rank sum scores of each of sub-domains will be recorded using Kruskal Wallis rank sum test among as well as the different clinical (LCL, MCL and DCL) types of lesions to check known-group construct validity. Wilcoxon-signed rank test will be used to compare mean difference of scores before initiation of treatment (Day 0) and at Day 90 (date of initial cure assessment).

Objective 4: To develop a severity scale for CL.

A round table CL expert discussion forum, informed by preliminary data, will be held, and consensus reached about potential items and scoring system for a severity scale. The prototype will be tested in individuals with CL at diagnosis.

Objective 5: To understand the lived experiences of individuals with CL.

Recordings will be transcribed and translated verbatim, and interviewer’s notes included for each interview. A codebook will be developed, and thematic analysis will be used. MAXQDA plus 2022 will be used to manage the in-depth interview data.

Objective 6: To assess household costs of CL, and health care provider, and societal costs of treating CL.

We will assess costs from a disaggregated societal perspective, examining costs to providers (MOH, donors), households, and society as a whole (i.e., costs to providers and household combined). Incremental costs to providers will be estimated relative to a hypothetical counterfactual in which no treatment for CL is provided. Incremental costs to households will be estimated relative to a hypothetical counterfactual in which the CL infection has not occurred. We will estimate total costs of diagnostic and treatment strategies and average costs, including cost per case diagnosed and cost per case resolved, disaggregated by site, type of CL, and patient characteristics (where possible), with appropriate statistical measures of variation. We will explore the degree of impoverishing and catastrophic expenditure for families of individuals with CL.

Descriptive statistics frequencies, percentages, means and medians will be used as appropriate to estimate costs. Further analysis will be conducted to investigate catastrophic out-of-pocket health expenditure and poverty impact using a linear probability model. STATA version 17 will be used for data analysis

Objective 7: To characterise how cell-mediated immune responses associate with treatment responses in CL due to L. aethiopica and (HLA) variation correlate with different clinical phenotypes.

Biopsies of affected skin and blood samples will be analysed to (i) evaluate secretion and quantification of selected host cytokines via the LuminexⓇ platform, (ii) evaluate the expression of innate/adaptive, and pro-and anti-inflammatory biomarkers and by evaluating the change in mRNA expression level of TLRs and pro- and anti-inflammatory cytokines before treatment, during treatment, and after completion of treatment.

Statistical analyses of laboratory data will be performed using R (version 4.4.1, the R Foundation for Statistical Computing). Nonparametric Mann-Whitney statistics will be used to compare different groups. Differences in cytokine and chemokine concentrations and mRNA expressions among CL types, treatment outcomes, time points, and other variables will be compared by the Kruskal-Wallis and t-tests. Fisher's exact test will be employed to determine the association between HLA types and CL types.